CHAPTER 2

LITERATURE REVIEW

7  

2. LITERATURE REVIEW

2.1 Solar dryer history

Drying is a simple process of removing excess water or moisture from a product

in order to reach the requirement of standard specification moisture content. Drying is

important especially to reduce the food product moisture content, as usually these have

much higher water content than the one that is suitable for long preservation. Reducing

moisture content of food product down to a certain level slows down the action of

enzymes, bacteria, yeasts and molds. Thus food can be stored and preserved for long

time without spoilage. Drying also is done with the aim of total removal of moisture

until food has no moisture at all. Dehydrated food, when ready to use, is re-watered and

almost regains its initial conditions. Drying is one of the most important post harvest

process for agriculture product. It can extend shelf life of the product, improve the

quality and reduce post-harvest losses due to waste. The transportation cost is also

reduced as the weight is less since the water is taken out from the product during the

drying process.

There are many methods available for drying process including hot air drying

using heater, dielectric drying (radio frequency or microwaves being absorbed inside the

material), freeze drying (drying method where the solvent is frozen prior to drying and

is then sublimed) and solar drying. Increase in the fuel price and awareness to protect

the environment had increase the potential usage of drying process using the solar

energy. Murthy (2009) defines the solar drying process as a process where the solar

radiation is used to evaporate the moisture present in the product. The solar energy is

used to heat large volume of air and this air is allowed to flow over the product to

remove and take away the moisture. Drying by solar energy is an economical procedure

for agricultural products, especially for medium to small agriculture industry, to prevent

excess of damage product after long storage. It is friendly to the environment. It used

8  

for domestic up to small size drying of crops, agricultural products and food product,

such as fruits, vegetables and herbs where its contribute significantly to the economy of

small agricultural communities and farms.

Traditionally, direct sun drying was performed by spreading the product on the

platform directly under the sun without any cover. However, according to Belessiotis

and Delyannis (2010), this method has many disadvantages including:

1) No scientific observations during long period of drying. The whole process depends

on the experience of unskilled labour.

2) No standard control on the final quality of the final dried product, it is just based on

observation and experience.

3) Very slow rate process, depending on the nature of the product and weather

condition.

4) The product is exposed directly to all kinds of weather changes, such as rain and

strong winds which can rot or destroy the material. Bad weather conditions on the other

hand facilitate growing of bacteria and molds.

5) They have very large qualitative and quantitative losses due to natural attack

conditions closely related to the open-air procedure such as dusting, rotting when

weather conditions are not favourable, attacks by insects, rodents, birds and other

unpredictable conditions.

In other words, the quality of finished product is inconsistent and difficult to

control. To overcome these problems, solar dryer was developed. Solar dryer is an

equipment which uses solar energy to heat up air and dry the food product. A solar

dryer minimizes almost all the problems faced during conventional sun drying method,

thus improving the quality of the dried product. According to Belessiotis and Delyannis

(2010) in comparison to conventional sun drying, the use of appropriate solar dryers

lead to a reduction of the drying time up to 50% and to a significant improvement of the

9  

product quality in terms of colour, texture and taste. Furthermore, contamination by

insects and micro organisms can be prevented. The storage losses can be reduced to a

minimum while the shelf life of the products can be increased significantly.

Several studies have been carried out to develope solar dryers for agriculture

products. Figure 2.1 below shows the classification of available solar dryers for

agriculture products based on the design of the system component and mode of

utilization of solar energy (Fudholi et al., 2010)

Figure 2.1: Classification of solar dryers and dryer modes.

(Fudholi et al., 2010)

10  

2.2 Design and development of greenhouse solar dryer

As shown in Figure 2.1 there are two types of solar dryer greenhouse, which are

natural circulation solar dryer greenhouse (passive dryer) and forced circulation solar

dryer greenhouse (active dryer). Compared to passive greenhouse dryers, in the active

solar dryer greenhouse the hot air was circulated by means of a ventilator. Naturally

ventilated greenhouse for drying applications have been reported in the past study and it

was reported that the dryer produced high quality dried food grades up to the desired

moisture content level (Sethi and Arora, 2009). Several design of solar dryer

greenhouse has been study from a simple small scale to a large scale solar dryer.

Koyuncu (2006) have designed, constructed and tested the performance of two different

types of natural circulation greenhouse type crop dryers (Figure 2.2a and Figure 2.2b).

He have developed a small scale (1 m x 1 m) greenhouse type solar dryer consist of

framework constructed from black coated metal bars, corrosion-resistant plastic mesh

and a black coated solar radiation absorber surface. The frameworks of the dryers were

clad with clear polyethylene sheet on the all sides. The cladding at rear side was

arranged to allow putting the moist products into the drying chamber or getting dried

product from there. The clear plastic cladding at the bottom edge of the front side and

rear side was also arranged to allow air to flow into the chamber, while the rectangular

stream at the top of the end served as the exit for the moist exhaust air. The results of

the study show that the greenhouse solar dryers increase the ambient air temperature by

5 to 9oC, and these dryers are 2 to 5 times more efficient than plastic mesh platform

type open sun dryer. The dryers with a drying air outlet chimney give better value of air

mass flow by increasing the air velocity. Ekechukwu and Norton (1996) have designed

and developed natural convection solar dryers which are suitable for the drying of most

crops. The design is a simplified design of the typical greenhouse type natural

11  

convection solar dryer (Figure 2.3). It consists of a cylindrical polyethylene-clad

vertical chamber, supported structurally by a steel framework and draped internally with

a selectively absorbing surface. They reported that performance of the dryer studied was

dependent largely on the variations in ambient temperature and relative humidity. The

results obtained from experimental solar chimneys in this study, if designed properly

could maintain chimney air temperatures consistently above the ambient temperature

which would enhance the desired buoyancy induced airflow through the chimney and

drying rate. Linear correlations have been obtained between the drying rate measured

experimentally and a group of ambient and crop parameters. Janjai et al. (2011) reported

that the large scale tunnel type solar dryer using polycarbonate cover (Figure 2.4) have

been tested and demonstrated potential of drying chilli, coffee and banana. The black

painted solar absorber surface raises the efficiencies of the dryers. In general, the solar

dryer offers much superior quality product compares to open sun drying

(a) (b)

Figure 2.2: Different types of natural circulation greenhouse type crop dryers.

(Koyuncu, 2006)

   

12  

Figure 2.3: Natural convection solar dryer greenhouse.

(Ekechukwu and Norton, 1996)

Figure 2.4: Large scale tunnel type solar dryer using polycarbonate cover.

(Janjai et al., 2011)

13  

2.3 CFD application in greenhouse development.

Computational fluid dynamics is a sophisticated design and analysis tool that

uses computers to simulate fluid flow, heat and mass transfer, phase change, chemical

reaction, mechanical movement, and solid and fluid interaction. The technique enables a

computational model of a physical system to be studied under many different design

constraints. CFD had the ability to efficiently develop spatial and temporal field

solutions of fluid pressure, temperature and velocity, and has proven its effectiveness in

system design and optimisation within many industries. The application of CFD in the

agricultural industry is becoming more important due to the above mentioned factor.

The versatility, accuracy and user-friendliness offered by CFD had led to its increased

take-up by the agricultural engineering community to study and analyse the indoor

climates of the greenhouse. This is reported by Norton et al. (2007) by the increase in

peer reviewed papers of CFD applications in agriculture buildings (Figure 2.5).

Figure 2.5: The number of published peer-reviewed publications of CFD applied to the

ventilation of agricultural buildings.

(Norton et al., 2007).

14  

2.3.1 Two dimensional and three dimensional CFD analyses.

Many studies around the world have been carried out to investigate the indoor

climate pattern of greenhouse structure using the CFD simulation. When CFD first used

to model airflow inside a room, it was assume that symmetric rooms with two

dimensional boundary conditions had a two dimensional airflow pattern. Molina-Aiz et

al. (2004) have done a study to analysed using computational fluid dynamics effect of

wind speed on the natural ventilation of an Almer´ıa-type greenhouse. He is using the

commercial program ANSYS/FLOTRAN v6.1 based on the finite elements method.

The experiment was carried out in an Almer´ıa-type greenhouse equipped with top and

side ventilation. The importance of roof ventilators for efficient ventilation in Almer´ıa-

type greenhouses was observed. The air temperature distribution shows a gradient from

the sidewalls towards the centre of the greenhouse due to the movement of the hot air

rising towards the roof vent, and a vertical gradient due to the movement of the air

above the surface of the ground absorbing solar energy at floor-level. Maximum air

velocity inside the greenhouse was reached near the side vents, with the lowest values

observed in the middle of the greenhouse. The velocity decrease produced in the

windward opening between the outside and inside of the greenhouse was 75 to 85% in

every case. The air velocity in the leeward area remained more or less constant around

0.3ms−1, as the result of the “chimney effect”. The model was verified by comparing the

numerical results with experimental data. The differences between values predicted by

the CFD models and those measured were from 0.0 to 0.36ms−1 for air velocities, and

from 0.1 to 2.1 ◦C for air temperatures. Although this study shows a good agreement

between predicted and measure data, later study found that many two dimensional CFD

investigation should be view with caution unless there is proof that these pattern would

occur in a physical situation because two-dimensional CFD predictions of such a large-

scale structure cannot be generalised (Norton et. al 2007). Observation made by Molina-

15  

Aiz et al. (2005) as quote by Norton et al. (2007) in a 2D study shows that ventilation

rate is reduced of around 88% when the span of the building was increased from scale 1

to 5. Large temperature gradients were also observed in the middle of the building under

all vent configurations, and insect screens were seen to greatly affect the ambient

temperature and velocity difference between indoor and outdoor environments.

The study using three dimensional CFD using Airpack 2.10 Fluent Inc. Software

was carried out by Pontikakos et al. (2006) to study the efficiency of natural ventilation

in a commercial twin-span greenhouse. In his study, three-dimensional patterns for

temperature and airspeed inside the greenhouse were generated, using specific boundary

conditions. The CFD simulation was applied to an empty twin-span greenhouse with a

floor area of 980 m2 with low density polyethylene cover and two side continuous

opening and one roof opening. The simulations assumed a sunny summer day with

different airspeeds (0.0, 1.0, 2.0 and 5.0 m/s) in three directions for each non zero

airspeed and three different temperature values for each airspeed (20.0, 25.0 and

30.0oC). The results showed that the external boundary temperature, wind directions

and airspeed are the crucial parameter on the pattern of the internal greenhouse

temperature. Campen and Bot (2003) also have used the three dimensions CFD (using

Fluent v.5.2 software) to study the ventilation of a Spanish ‘parral’ greenhouse. The

simulation was verified by experimental result using tracer gas measurement. The result

shows the simulation data and the experiment data was resembled within 15%. From his

study he concludes that a three dimension CFD model was able to determine the

greenhouse specific ventilation characteristic. The CFD calculation also indicates that

the ventilation rate is largely depend on wind direction and the wind speed was

correlated linearly with ventilation rate without the buoyancy effect.

16  

2.3.2 CFD analysis for mono span greenhouse with pitched roof design.

The usage of CFD to investigate the mono-span type greenhouse with pitched

roof design has been carried out by Shklyar and Arbel (2004). They study the effect of

vent angles and wind direction on wind induced ventilation in an isothermal, pitched

roof single span type of greenhouse. They found that by changing the vent angle from

20o to 40o can doubled the ventilation rate. They also found that the change in wind

direction from 45o to 90o had more effect to ventilation rate compare to wind direction

from 0o to 45o. Which mean ventilation rate induced by perpendicular wind was almost

five times greater than when wind is parallel to the structure. Another study on pitched

roof greenhouse was conduct by Campen (2005) to study the climate distribution inside

the greenhouse under different wind speed and direction, various porous screen and

different structure configuration. The result of the study shows that the resistant from a

screen net gives more influence on the ventilation rate compare to wind direction. At

low wind speed, 0.5 ms-1, a greenhouse with roof and sidewall ventilation gave a large

temperature increase inside the building. This was caused by both buoyancy and wind

force counteracting one another. The temperature inside the building is increasing when

the greenhouse design configuration was lengthened from 12m to 36m.

17  

2.4 Governing equation

2.4.1 Navier-Stokes equation

CFD programs numerically solve Navier-Stokes and energy equation. The

Navier-Stokes equations are the basic governing equations for a viscous, heat

conducting fluid. It is a vector equation obtained by applying Newton's Law of Motion

to a fluid element and is also called the momentum equation. It is supplemented by the

mass conservation equation, also called continuity equation and the energy equation.

The term a Navier-Stokes equation is used to refer to all of these equations. These

equations constitute a micro model of fluid motion, and required time consuming

iterative technique to solve them. In ANSYS CFX the instantaneous equation of mass,

momentum and energy conservation can be written as follows (CFX-Solver theory

guide, 2009):

a) The continuity equation

. (2.1)

b) The momentum equation

. . (2.2)

Where the stress tensor, τ is related to the strain rate by

. (2.3)

c) The total energy equation

. . . . . (2.4)

Where is the total enthalpy, related to the static enthalpy h (Top) by:

(2.5)

The term . . represents the work due to viscous stresses and is called the viscous

work term

18  

The term . represent the work due to external momentum sources and it’s currently

neglected.

Besides Navier-Stokes equations, account must also be taken of the additional

processes that may influence the dynamics of ventilation system. The governing

equation may need to be modified with additional physical models or assumption to

fully represent the physical situation. This may includes turbulence and porous media

model, and models that describe occupant inside the building.

2.4.2 Turbulence model

Turbulence motions are usually associates with ventilation primarily due to high

flow rates and heat transfer interaction involved in the flow field. Currently there are

many turbulence models available and many studies have been conducted to validate

these turbulence model. One of best performing turbulence model that being used in the

modelling of agriculture buildings and application is the standard k-ε model (Norton,

2007). This model introduces two variables which are the turbulence energy, k and its

dissipation rate, ε. The variable k corresponds to the turbulence velocity of the mixing

length model and the variable ε corresponds to mixing scale. The model adds two

transport equations corresponding to the two new variables to the usual transport

equations describing the flow. In this way no mixing length needs to be defined and

therefore complex flow governed by elliptic equation, such as recirculation flows can be

solve. The standard k-ε model equation is written as follow:

19  

For turbulent kinetic energy, k

ε (2.6)

For dissipation, ε

(2.7)

Where,

μt is turbulent viscosity, Pk is production of k, S is the modulus of the mean rate of strain

tensor, Pb is effect of buoyancy, C1ε = 1.44, C2ε = 1.92, Cμ=0.09, σk = 1.0 and σε = 1.3.

One important weakness of the standard k-ε model is that it does assume

equilibrium. Meanings that one’s turbulence energy is generated at the small wave

number end of the spectrum (large eddies), it is equally distribute to the whole

spectrum. Generally this is not the case because the transfer of energy from the large

eddies, where turbulence is produce to the small eddies and turbulence dissipation

occurs is not automatic. A considerable length of time intervenes between the

production and the dissipation of turbulence. Moreover this process can be affected by

the interaction with obstruction and walls (Mistiiotis et. al, 1997).

To overcome the above mention problem, two scale k-ε turbulence model have

been introduced. Chen-Kim model and the renormalized group (RNG) k-ε model is the

two scale k-ε turbulence model that introduce to model air movement within and around

buildings. Chen-Kim model improves the dynamic response of the equation for k by

introducing a second time scale, k/p where p is the volumetric production rate of k.

Furthermore several of the standard model coefficients are adjust so that the models

maintain good agreement with experiment data on classical turbulent shear layer flows.

In RNG model, large scale is described by renormalized group method, where the effect

of the small scale is represented by modified transport coefficient. In two dimension

20  

wind induced ventilation, Mistriotis et al. (1997) showed that better qualitative

agreement with experiment observed flow patterns can be achieved with two scale k-ε

model than with standard k-ε model. Later Roy and Boulard (2005) and Brugger et al.

(2005) as quote by Norton (2007) also found a difference in ventilation rate predicted by

two scale k-ε model and standard k-ε model. However when heat transfer is coupled

with the field flow all k-ε model seem to perform similarly and shown similar

agreement with experimental data.

Another new turbulence model is the Shear Stress Transport (SST) which

combine the k-ε and k-ω models using a blending function was found to predict airflow

in good agreement with experimental data not only in the near wall region of the flow

but also in the free stream. Toma’s et al (2007) suggests that this model should be

considered when buoyancy force plays a major role in driving the flow within a

ventilated space. Another turbulence model that has been used in previous study is the

Reynolds stress closure models (RSM). Reynolds stress closure models (RSM) have

exhibited far superior predictions for flows in confined rooms where adverse pressure

gradients occur. These models have recently been shown to enhance the CFD

predictions of primary and secondary flow patterns in empty-isothermal rooms and in

loaded rooms with heat transfer (Toma’s et al, 2007). However according to Toma’s et

al (2007) the weakness of RSM is extra computational time and memory required in

solving the flow regime, alongside the difficulties in attaining good convergence

behaviour.

21  

Figure 2.6: Influence of the turbulence model on the flow pattern at the symmetry plane

of a confined flow.

(Moureh and Flick, 2005 as quote by Norton et al, 2007)

2.4.3 Heat transfer model.

In this study the main factor that influences the internal climate of the solar dryer

greenhouse is the sun radiation. Heat transfer in a fluid domain is governed by the

energy transport equation. In this study the radiation effect is significant compare to the

convective and conductive heat transfer rates, therefore to account for radiation,

Radiative Intensity Transport Equations (RTEs) are solved. Several radiation models are

available in ANSYS CFX which provides approximate solutions to the RTE are

(ANSYS CFX-Solver theory guide, 2009):

1) Rosseland Model (Diffusion Approximation Model)

2) P-1 Model (Gibb’s Model/Spherical Harmonics Model)

3) Discrete Transfer Model (DTM) (Shah Model)

4) Monte Carlo Model

22  

Each radiation model has its assumptions, limitations, and benefits. Rosseland

model is a simplification of RTE for the case of optically thick media. Its introduce new

diffusion term into the energy transport equation with a strongly temperature dependant

diffusion coefficient. The P1 model is also a simplification of RTE. It assume that the

radiation intensity is isotropic or direction independent at a given location in space.

DTM assume that the scattering of the radiation is isotropic. The Monte Carlo Model

assumes that the intensity is proportional to the differential angular flux of photon. The

optical thickness should be determined before choosing a radiation model. For thin

optical meaning that the fluid is transparent to the radiation at wavelengths where the

heat transfer occurs and the radiation only interacts with the boundaries of the domain.

While thick optical means that the fluid absorbs and re-emits the radiation. For optically

thick media the P1 model is a good choice that gives reasonable accuracy without too

much computational effort. Monte Carlo and DTM model also can be used for optically

thick media, but the P1 model uses far less computational resource. For optically thin

media, the Monte Carlo or DTM may be used but for models with long or thin

geometries DTM can be less accurate. While Monte Carlo uses the most computational

resources compared to DTM.

2.4.4 Modelling for buoyancy.

The ventilation in the solar dryer greenhouse in this study is natural ventilation. In

natural ventilation, natural convection occurs when temperature differences in the air,

resulting in density variations. This is called the buoyancy driven flow. Buoyancy

driven flow is generated in the air when part of the air is heated or cooled by a surface.

This causes motion in the air as the warm air rises and the cool air is then moved to the

surface and will become heated. Climatic variables inside greenhouse are a function of

23  

varying flow properties caused by heating and cooling of the air. There are two main

methods of modelling the density variation that occur due to buoyancy. First is the

Boussinesq approximation. This has been used successfully in many housing

application. The density variation according to Boussinesq approximation is as below

(Toma’s et al, 2007)

1 (2.8)

Where is the thermal expansivity:

| (2.9)

And Tref is the buoyancy reference temperature.

The assumption involved in this approximation involved:

a) Density differentials in the flow are only required in the buoyancy term of the

momentum equation.

b) There is linear relationship between temperature and density, with all other extensive

fluid properties being constant.

c) The temperature different in the flow field is less than 30oC.

The main drawbacks of this approximation is that its only consider dry air as the

fluid medium, while in actual condition most climatic flow will involved a mixture of

dry air and moisture. The extended version of Boussinesq approximation was derived

by Gan (1994) which describe the density of moist air as a function of temperature and

moisture concentration. The extended version can be expressed as

(2.10)

Where C is water concentration and

and However despite the enhance relation,

according to Toma’s et al (2007) this model is not commonly used in the literature. For

condition where there is a large temperature different, the Boussinesq approximation is

24  

not sufficiently accurate. Therefore another method is used where t is done by treating

the air as an ideal gas and expressing the density different by means of the ideal gas

equation:

(2.11)

In this equation the density of the fluid is dependent on temperature and composition

but not pressure. Although this equation may provide an accurate description of the

density variation within the flow regime, it has been found to have impact on

convergence behaviour of CFD solution.